Author Topic: Ripple current rating of capacitors in parallel  (Read 288 times)

0 Members and 1 Guest are viewing this topic.

Offline multime

  • Contributor
  • Posts: 23
Ripple current rating of capacitors in parallel
« on: July 15, 2021, 02:30:34 pm »
I am just in the process of constructing a bench power supply. It is based on a old Vellman kit which i think was labelled as "0-30v 10amp laboratory power supply" I had built this same power supply many years back and was satisfied with its performance just some annoyance getting the displays to work properly. I am now selecting a suitable smoothing capacitor/s and was looking to use a 22000uf unit ( had a 100000uf in the last build,  might have been overkill) due to cost and availability in the UK i am thinking of going with two x 11000uf in parallel which are not a bad price. I noticed when looking at there spec that they allow a ripple current of 9.4amps obviously slightly short of the 10amp in the Vellman title and i was wondering as they will be in parallel how would this effect the ripple current for each of them.

 

Offline Benta

  • Super Contributor
  • ***
  • Posts: 3609
  • Country: de
Re: Ripple current rating of capacitors in parallel
« Reply #1 on: July 15, 2021, 02:51:58 pm »
First, ripple current has little connection to the output current of your power supply. They're approximately proportional, but not equal.
That being said, they'll share the ripple current well enough for it to be no issue.

 

Offline multime

  • Contributor
  • Posts: 23
Re: Ripple current rating of capacitors in parallel
« Reply #2 on: July 15, 2021, 03:05:06 pm »
Thanks for the quick reply and good knowing no problems with my choice.

 

Online TimFox

  • Super Contributor
  • ***
  • Posts: 3554
  • Country: us
  • Retired, now restoring antique test equipment
Re: Ripple current rating of capacitors in parallel
« Reply #3 on: July 15, 2021, 03:40:17 pm »
To a reasonable approximation, the rms ripple current through a filter capacitor at 100/120 Hz is the rms ripple voltage divided by the reactance at 100 or 120 Hz. 
The ESR of the capacitor is important to determine its maximum ripple current specification, but has little effect on the current (so long as ESR << XC, the reactance).  At 100 Hz, the reactance of 11 mF is 0.14 ohms, while a good aluminum capacitor of that value has an ESR of about 0.03 ohms. 
Interestingly, at no DC load current, the ripple across the capacitor is very small (theoretically zero), but the ripple voltage increases as you increase the DC load current, roughly proportional to the current.
Since the voltage across each capacitor is equal (parallel connection), the ripple currents will be equal (to within the capacitance tolerance).
For a detailed calculation of traditional power supply rectifiers and filters, I refer you to "PSUD2", free software written by Duncan, https://www.softpedia.com/get/Science-CAD/PSU-Designer-II.shtml
(Note that the transformers are specified by no-load voltage and effective secondary resistance, which can be calculated from the catalog output voltage at specified current, and a specified or estimated efficiency.)
 

Online Kleinstein

  • Super Contributor
  • ***
  • Posts: 10033
  • Country: de
Re: Ripple current rating of capacitors in parallel
« Reply #4 on: July 15, 2021, 04:55:25 pm »
Many of the low cost kits claim a maximum current higher than they should actually be used at. It starts with the transformer rating that should be something like 1.6-1.8 times higher than the Dc current used. So a 10 A transformer is good for some 6 A, unless one has extra power factor correction at the rectifier.
With a constant current draw the ripple current should be a little less than the average of the ADC and DC current. So a ripple current about as high as the DC current is not a bad approximation.
 

Offline multime

  • Contributor
  • Posts: 23
Re: Ripple current rating of capacitors in parallel
« Reply #5 on: July 15, 2021, 09:27:25 pm »
TimFox, thanks for the info. Took a look at the download link and attempted to download but win10 had other ideas and blocked it and of course it never gives you a choice of download the file anyway and stop trying to be my nanny. will take a look tomorrow.

Kleinstein, handy rule of thumb but what is "ADC"

 

Online TimFox

  • Super Contributor
  • ***
  • Posts: 3554
  • Country: us
  • Retired, now restoring antique test equipment
Re: Ripple current rating of capacitors in parallel
« Reply #6 on: July 15, 2021, 09:32:14 pm »
Try these links for the software:  https://groups.io/g/duncanampspsud  or  http://www.duncanamps.com/psud2/
I have no trouble running the software on Windows 10, but I don't remember whence I downloaded it.
 
The following users thanked this post: Just_another_Dave


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf